Tensorflow Neuron (“tensorflow-neuronx”) Setup on Amazon Linux 2 with DLAMI Tensorflow
Contents
This document is relevant for: Inf2
, Trn1
, Trn1n
Tensorflow Neuron (“tensorflow-neuronx”) Setup on Amazon Linux 2 with DLAMI Tensorflow#
Table of contents
Note
Please make sure to upgrade
from python 3.7
to python 3.8
to use Neuron SDK on Amazon Linux 2
. Starting from Neuron Release 2.13
, python 3.7
is no longer supported as mentioned here.
Also, we do not have support for torch-neuronx 2.1.2
on Amaznon Linux 2.
Get Started with Latest Release of Neuron Tensorflow (tensorflow-neuronx
)#
This section provide links that will assist you to quickly start with a fresh installation of TensorFlow Setup Guide for Inf2 & Trn1.
Launch the Instance
Please follow the instructions at launch an Amazon EC2 Instance to launch an instance. When choosing the instance type at the EC2 console. please make sure to select the correct instance type.
To get more information about instances sizes and pricing see: Trn1 web page, Inf2 web page
Check for the latest version of the Deep Learning AMI Neuron TensorFlow 2.10 and copy the AMI name that starts with “Deep Learning AMI Neuron TensorFlow 2.10 (Amazon Linux 2) <latest_date>” from “AMI Name:” section
Search for the copied AMI name in the AMI Search , you should see an exact matching AMI with the AMI name in Community AMIs. Select the AMI and use it to launch the instance.
When launching a Trn1, please adjust your primary EBS volume size to a minimum of 512GB.
After launching the instance, follow the instructions in Connect to your instance to connect to the instance
Update Neuron Drivers
# Update OS packages
sudo yum update -y
# Update OS headers
sudo yum install kernel-devel-$(uname -r) kernel-headers-$(uname -r) -y
# Install git
sudo yum install git -y
# upgrade Neuron Driver
sudo yum update aws-neuronx-dkms-2* -y
# Update Neuron Runtime
sudo yum update aws-neuronx-collectives-2.* -y
sudo yum update aws-neuronx-runtime-lib-2.* -y
# Update Neuron Tools
sudo yum update aws-neuronx-tools-2.* -y
# Add PATH
export PATH=/opt/aws/neuron/bin:$PATH
Get Started With Tensorflow DLAMI
# Activate Python venv
source /opt/aws_neuron_venv_tensorflow/bin/activate
# Install Jupyter notebook kernel
pip install ipykernel
python3.8 -m ipykernel install --user --name aws_neuron_venv_tensorflow --display-name "Python (tensorflow-neuronx)"
pip install jupyter notebook
pip install environment_kernels
# Set pip repository pointing to the Neuron repository
python -m pip config set global.extra-index-url https://pip.repos.neuron.amazonaws.com
# Update Neuron Compiler and Framework
python -m pip install --upgrade neuronx-cc==2.* tensorflow-neuronx
Update to latest TensorFlow Neuron (tensorflow-neuronx
)#
If you already have a previous Neuron release installed, this section provide links that will assist you to update to latest Neuron release.
Note
Instructions in this page only apply to setting up Neuron components on Linux host running Ubuntu or Amazon Linux AMI.
When launching a Trn1, please adjust your primary EBS volume size to a minimum of 512GB.
# Activate Python venv
source /opt/aws_neuron_venv_tensorflow/bin/activate
# Install Jupyter notebook kernel
pip install ipykernel
python3.8 -m ipykernel install --user --name aws_neuron_venv_tensorflow --display-name "Python (tensorflow-neuronx)"
pip install jupyter notebook
pip install environment_kernels
# Set pip repository pointing to the Neuron repository
python -m pip config set global.extra-index-url https://pip.repos.neuron.amazonaws.com
# Update Neuron Compiler and Framework
python -m pip install --upgrade neuronx-cc==2.* tensorflow-neuronx
Note
Instructions in this page only apply to setting up Neuron components on Linux host running Ubuntu or Amazon Linux AMI.
When launching a Trn1, please adjust your primary EBS volume size to a minimum of 512GB.
# Activate Python venv
source /opt/aws_neuron_venv_tensorflow/bin/activate
# Install Jupyter notebook kernel
pip install ipykernel
python3.8 -m ipykernel install --user --name aws_neuron_venv_tensorflow --display-name "Python (tensorflow-neuronx)"
pip install jupyter notebook
pip install environment_kernels
# Set pip repository pointing to the Neuron repository
python -m pip config set global.extra-index-url https://pip.repos.neuron.amazonaws.com
# Update Neuron Compiler and Framework
python -m pip install --upgrade neuronx-cc==2.* tensorflow-neuronx==2.9.3.2.1.0
Note
Instructions in this page only apply to setting up Neuron components on Linux host running Ubuntu or Amazon Linux AMI.
When launching a Trn1, please adjust your primary EBS volume size to a minimum of 512GB.
# Activate Python venv
source /opt/aws_neuron_venv_tensorflow/bin/activate
# Install Jupyter notebook kernel
pip install ipykernel
python3.8 -m ipykernel install --user --name aws_neuron_venv_tensorflow --display-name "Python (tensorflow-neuronx)"
pip install jupyter notebook
pip install environment_kernels
# Set pip repository pointing to the Neuron repository
python -m pip config set global.extra-index-url https://pip.repos.neuron.amazonaws.com
# Update Neuron Compiler and Framework
python -m pip install --upgrade neuronx-cc==2.* tensorflow-neuronx==2.8.4.2.1.0
Install Previous TensorFlow Neuron Releases for Amazon Linux (tensorflow-neuronx
)#
This section will assist you to install previous Neuron releases.
# Install Python venv
sudo yum install -y python3.8-venv gcc-c++
# Create Python venv
python3.8 -m venv aws_neuron_venv_tensorflow
# Activate Python venv
source aws_neuron_venv_tensorflow/bin/activate
python -m pip install -U pip
# Install Jupyter notebook kernel
pip install ipykernel
python3.8 -m ipykernel install --user --name aws_neuron_venv_tensorflow --display-name "Python (tensorflow-neuronx)"
pip install jupyter notebook
pip install environment_kernels
# Set pip repository pointing to the Neuron repository
python -m pip config set global.extra-index-url https://pip.repos.neuron.amazonaws.com
# Install wget, awscli
python -m pip install wget
python -m pip install awscli
# Install Neuron Compiler and Framework
python -m pip install neuronx-cc==2.12.68.0 tensorflow-neuronx==2.10.1.2.1.0
# Install Python venv
sudo yum install -y python3.8-venv gcc-c++
# Create Python venv
python3.8 -m venv aws_neuron_venv_tensorflow
# Activate Python venv
source aws_neuron_venv_tensorflow/bin/activate
python -m pip install -U pip
# Install Jupyter notebook kernel
pip install ipykernel
python3.8 -m ipykernel install --user --name aws_neuron_venv_tensorflow --display-name "Python (tensorflow-neuronx)"
pip install jupyter notebook
pip install environment_kernels
# Set pip repository pointing to the Neuron repository
python -m pip config set global.extra-index-url https://pip.repos.neuron.amazonaws.com
# Install wget, awscli
python -m pip install wget
python -m pip install awscli
# Install Neuron Compiler and Framework
python -m pip install neuronx-cc==2.12.54.0 tensorflow-neuronx==2.10.1.2.1.0
# Install Python venv
sudo yum install -y python3.8-venv gcc-c++
# Create Python venv
python3.8 -m venv aws_neuron_venv_tensorflow
# Activate Python venv
source aws_neuron_venv_tensorflow/bin/activate
python -m pip install -U pip
# Install Jupyter notebook kernel
pip install ipykernel
python3.8 -m ipykernel install --user --name aws_neuron_venv_tensorflow --display-name "Python (tensorflow-neuronx)"
pip install jupyter notebook
pip install environment_kernels
# Set pip repository pointing to the Neuron repository
python -m pip config set global.extra-index-url https://pip.repos.neuron.amazonaws.com
# Install wget, awscli
python -m pip install wget
python -m pip install awscli
# Install Neuron Compiler and Framework
python -m pip install neuronx-cc==2.11.0.34 tensorflow-neuronx==2.10.1.2.1.0
This document is relevant for: Inf2
, Trn1
, Trn1n